The present invention relates to an apparatus for reading and analyzing an emitted light (fluorescent light) pattern emitted from a sample located on a plane. More particularly, it relates to a method and an apparatus by which living body samples, such as a DNA, a RNA and an oligonucleotide marked by a fluorescent material, are trapped at a plurality of positions on a substrate so as to read an emitted light pattern from the fluorescent marker with a high-resolution, at a high-speed and with a high-sensitivity.
Analysis technologies for analyzing the DNA and proteins are important in fields of medicine and biology including gene analysis and gene diagnosis. In particular, in recent years, attention has been focused on a method and an apparatus by which, using a DNA probe array (which is called in other way by various names such as oligochip, DNA chip or biochip, but hereinafter the DNA probe array is employed as the generic name), a variety types of DNA sequence information and gene information are inspected and analyzed simultaneously from one specimen to be inspected. The DNA probe array is implemented in the following way: A substrate such as a glass is used and separated into a plurality of (millions to tens of millions) areas, and target (usually, different types of) DNA probes are fixed onto the respective areas so as to form the respective areas into microscopic reaction areas, thus implementing the array. The reaction of the specimen with the array allows a target DNA in the specimen to be trapped in a state of being hybridized with the fixed DNA probes. Moreover, the target DNA is coupled with a probe such as a fluorescent probe, thereby making it possible to measure the coupled state (the position, i.e., the hybridized sequence) and its quantity with the use of the fluorescence intensity and so on. This permits the array to be utilized in the gene diagnosis and the sequence.
A microscope-like apparatus (confocal fluorescent microscope), which is usually called a scanner, is used in order to read the fluorescence intensity emitted from the fluorescent marker on the target DNA trapped in the respective reaction areas of the DNA probe array (for example, refer to JP-A-11-315095). In this apparatus, the array is irradiated with a excitation light such as a laser light which has been modified into one microscopic spot-shaped configuration, and fluorescent lights generated are separated from the excitation light using a spectroscopic device such as an interference filter, then detecting the fluorescence intensity using a photo detector such as a photomultiplier tube. At that time, the excitation light is waved using a galvanometer mirror so as to scan, in a 2-dimensional manner, the microscopic spot formed on the array. Otherwise, the position of the microscopic spot is fixed, then scanning the array in the 2-dimensional manner. These scans make it possible to recognize the fluorescence intensity distribution of the entire array, i.e., the degree of the coupling toward each DNA probe.
When trying to apply the DNA inspection to a living body inspection such as a conventional blood inspection, it becomes absolutely required to execute the inspection toward a large number of living specimens at a high-speed. In the conventional technologies, however, it takes so much inspection time to irradiate one point with the excitation light spot, which is equivalent to a necessary inspecting resolution of the DNA chip, and to detect the obtained fluorescent lights one after another. This is concerned with the fact that it is impossible to shorten unlimitedly a time needed to detect this one point. Namely, this is attributed to the fact that the time ΔtL, which is needed up to the finishing of the fluorescent light generation after the irradiation with the excitation light, is equal to substantially 10 ns. Transferring the detection to the next detection point without waiting for the finishing of the fluorescent light generation makes the detection itself impossible.
Also, it is required to perform the detection with a high-sensitivity that corresponds up to a state where there exist several fluorescent molecules within the spot light size equivalent to the above-described necessary detecting resolution. However, all of the generated fluorescent lights are not detected. Namely, a light utilization efficiency of the detecting optical system and a quantum efficiency of the photomultiplier tube used for the optical detection are not equal to 100%. Moreover, the following values are small: An efficiency with which the excitation light is absorbed in the fluorescent material and a probability with which the absorbed excitation light is converted into the fluorescent light. This requires that the detection be executed by spending a time that is at least tens to hundreds times longer than ΔtL. Furthermore, the more this time is lengthened, the higher the accuracy of a weak light detection becomes which is almost the same as the photon counting.
Also, in order to accomplish such a high-speed characteristic at a practical level, the following conditions are required: Eliminating or reducing influences of foreign substances that are composed of various types of proteins mixed into the DNA chip, or maintaining all the time a focal point of the detecting system onto the inspection plane on which the fluorescent material-added target has been hybridized. Also, in some cases, it becomes necessary to perform the detection with reference to a plurality of fluorescent lights at a high-speed.
When forming multi-spot lights from a bean emitted from a laser light-source, if the emitted light is wished to be modified into the multi-spot lights without being wasted, i.e., for example, when trying to form 50 multi-spot lights that are 0.04 mm in diameter and are arranged in a line with a pitch of 0.4 mm, the emitted laser beam is formed so that the beam becomes an ellipse-shaped configuration the longitudinal-to-transverse ratio of which is equal to 1 to 50. Irradiating, with the ellipse-shaped beam, a microlens array where microlenses are arranged with a pitch of 0.4 mm makes it possible to obtain the 50 multi-spot lights about 0.04 mm in diameter and having the pitch of 0.4 mm, if a focal length of the microlenses is equal to 60 mm, for example. Since the intensity distribution of the laser beam is Gaussian distribution, in the arrangement of the 50 multi-spot lights formed in the method as described above, spots situated in proximity to the center of the arrangement become lighter and spots situated in proximity to the periphery thereof become darker. Trying to equate as much as possible the intensities of the spots at the center with those of the spots on the periphery requires the following: A spread area of the incoming ellipse-shaped beam should be made large enough as compared with the width of the 50 microlenses. As a result, half or more of the laser beam energy emitted from the laser light-source turns out to be wasted.
In the DNA probe array, the number of the types of the fixed probes is increased and thus its density is becoming more and more heightened. On account of this, an apparatus for measuring the fluorescent marker distribution of the DNA probe array is expected to operate at the high-speed and with the high-sensitivity in addition to implementing the high-resolution characteristic. In the above-described apparatus, however, no sufficient consideration has been given to this point. Namely, it is required to make the spot diameter of the excitation light more microscopic in order to implement the high-resolution characteristic, whereas the scanning time is increased at that case (when an exposure time for each spot is identical). Moreover, making the spot diameter smaller lessens the number of the fluorescent molecules existing within the spot diameter, thereby lowering the detection sensitivity. Also, in order to implement the high-speed characteristic, it is required to increase the scanning rate or to make larger the spot diameter of the excitation light. Increasing the scanning rate, however, shortens the excitation time for the fluorescent material, thereby lessening the fluorescence intensity or increasing a circuit noise to lower the detection sensitivity. Also, in order to implement the high-sensitivity characteristic, it is required to make the spot diameter larger and to delay the scanning rate. Namely, implementing the high-resolution characteristic, implementing the high-speed characteristic and implementing the high-sensitivity characteristic are contradicted to each other, and accordingly it was difficult to accomplish them.
Also, there has been proposed an apparatus by which a plurality of microscopic spots are formed on a sample surface and lights from the plurality of microscopic spots are detected simultaneously (for example, refer to JP-A-11-118446). In this configuration, a light-flux from an excitation light-source is enlarged and a 2-dimensional microlens array is located in the enlarged light-flux so as to form the plurality of irradiation spots, then projecting the spots onto the sample surface. In this case, however, especially when a laser light is used as the excitation light, because of the Gaussian distribution characteristic of the laser beam, the light intensities of spots in proximity to the center are intense and the light intensities of the spots become lower as their positions come nearer to the periphery. Accordingly, it is difficult to equate all the light intensities of the plurality of irradiation spots. This problem is partly solved by enlarging the excitation light up to a light-flux the diameter of which is large enough as compared with the width of the 2-dimensional microlens array. In this case, however, the light utilization efficiency is no good and the light intensities are decreased, thus resulting in such a problem that the detection sensitivity is deteriorated.
In a method of focusing a one-beam laser light onto a target to be inspected, the detection is executed on a one-pixel by one-pixel basis in a time-sequence manner. As a result, when the fluorescent light is extremely weak, it takes a time to execute the detection equivalent to the one-pixel. Consequently, it takes a considerable amount of time to detect the entire 2-dimensional image. Even if an excitation light of the one-beam laser is strengthened in order to detect the weak fluorescent light emitting target at a high-speed, when trying to detect, for example, 6000×6000 pixels in 1 minute, the detecting time per pixel turns out to become 1 to 2 μsec (microseconds). Executing the detection in such a short time makes it difficult to detect the fluorescent light detection intensity in a broad dynamic range (for example, 216). Also, since a normal fluorescent material emits the fluorescent light with a delay of 10−9 to 10−5 sec after receiving the excitation light, the above-described per-pixel fluorescent light detecting time of 1 to 2 sec prevents the fluorescent light detection from being fully accomplished. Meanwhile, in a method using Nippou disk, the light is utilized effectively only by the amount of a ratio of a microscopic aperture in a uniform light. This condition requires a considerable amount of time to be spent to detect the weak fluorescent light.